crushing05.github.io

I am a population ecologist with an emphasis on quantifying demographic processes and their influences on individuals, populations, and species. My research focuses primarily on the ecology and conservation of animal migration, with projects falling within three major themes: 1) Development of full-annual-cycle models to quantify seasonal vital rates and their influence on population dynamics; 2) Demographic consequences of climate change and incorporating climate change into on conservation planning for migratory birds; and 3) Quantifying the influence of long-distance dispersal on individuals and populations. To address these issues, my research combines observational and experimental fieldwork with quantitative modeling and stable isotope analysis.

Below are brief descriptions of past, current, and future research within each of these themes:

Development of full-annual-cycle models to quantify seasonal vital rates and their influence on population dynamics

Historically, efforts to identify limiting factors for migratory species have been hindered by the scale of annual migration movements, missing information about the geographic linkages between breeding and non-breeding populations (i.e., migratory connectivity), and the lack of analytical methods for linking population dynamics to demographic processes operating across the entire annual cycle. A major theme of my research is developing and parameterizing full-annual-cycle population models to overcome these challenges.

Much of my research on full-annual-cycle models has focused on a declining migratory bird, the Wood Thrush (Hylocichla mustelina). Like many migratory birds, Wood Thrush populations have suffered large population declines in recent decades but the environmental and demographic drivers of these declines remain poorly understood. Spatial variation in migratory connectivity (Rushing et al., 2014; Rushing et al., 2017a) and the rate of decline (Rushing et al., 2015a) suggest that regional populations may experience distinct threats during the breeding and non-breeding periods. To quantify the contribution of full-annual-cycle environmental processes to regional population declines, colleagues and I developed a hierarchical model to link annual variation in breeding abundance to forest loss and climate change on the breeding and winter grounds (Rushing et al., 2016a). Results from this analysis suggested that regional populations in the core of the breeding range are likely limited by winter habitat but also that the steepest regional declines have likely been driven by breeding habitat loss. In contrast, abundance in populations at the periphery of the range was more strongly influenced by winter climate. Collectively, these results indicate that limiting factors are not uniform across space and suggest that a one-size-fits-all approach to management may not be effective.

Ultimately, understanding populations dynamics and devising effective management strategies requires identifing the seasonal demographic processes that limit population growth. To address that issue, I developed a full-annual-cycle integrated population model (IPM) to link changes in breeding abundance to reproduction, immigration and seasonal survival. By integrating summer and winter capture-recapture data collected from linked populations, this model is able to parse annual survival into its seasonal components (summer, autumn, winter, and spring). Furthermore, by combining data on population size, survival, and reproduction, this model provides a full accounting of the demographic contributions to population dynamics. Application of the model to our Wood Thrush system provided the first separate estimates of spring and fall migration survival for a migratory songbird ([Rushing et al. 2017b]) and demonstrated that survival during spring migration is the largest demographic driver of population dynamics in this system.

Currently, I am working on a number of projects that use and expand these full-annual-cycle frameworks to answer questions about the ecology and conservation of migratory birds. These projects include: using simulated data to investigate the estimability of latent migration survival rates from integrated capture-recapture models; identifying the environmental drivers of migration survival in several migratory wood warblers; quantifying the effects of winter density-dependence on spring migration survival and population limitation of American Redstarts (Setophaga ruticilla); and identifying potential limiting factors for Rusty Blackbirds (Euphagus carolinus), the fastest declining songbird in North America.

Recently I have started a project with colleagues from the Smithsonian and Cornell focused on the southeastern Painted Bunting (Passerina ciris). The first phase of this project, initiated in summer 2017, involved deployment of light-level geolocators to quantify the winter distribution and migratory connectivity of this population. Future work will harness several large citizen-science programs to collect demographic monitoring data from linked summer and winter populations and develop a full-annual-cycle model to identify management actions to address recent populations declines.

Improving conservation outcomes through demographic modeling

Across the globe, migratory species are showing widespread, sustained population declines. Conservation actions to reverse these declines will only be successful if we can identify and target the primary threats driving these declines. Because population limitation can occur at any time during the annual cycle (i.e. breeding, non-breeding, and migration), conservation of migratory species requires linking seasonal, demographic and environmental processes and connecting demographic modeling to conservation planning efforts. The recent development of integrated population models (IPMs), which combine multiple sources of demographic monitoring data to estimate changes in population size as a function of vital rates, provides a powerful framework for advancing both ecological theory and conservation planning.

A major part of my current research is developing full-annual-cycle IPMs to link the population dynamics of migratory species to demographic processes, and ultimately environmental drivers, operating during the breeding, winter, and migratory periods. This framework incorporates multiple sources of data within a unified Bayesian modeling framework to jointly estimate seasonal vital rates and connect changes in population size to specific demographic processes. These models provide a general framework for modeling population processes of migratory species across the annual cycle and, for the first time in migratory songbirds, provide survival estimates during spring and fall migration. Application of this framework to Wood Thrush has demonstrated that survival during spring migration is the primary driver of population dynamics in this system but also that the relative importance of breeding vs. non-breeding vital rates varies as a function of habitat quality ([Rushing et al., 2017]). As part of the International Wood Thrush Conservation Alliance, I am also working with academic, government, and NGO partners to incorporate my modeling results into a full-annual-cycle conservation plan to guide future Wood Thrush management actions.

In addition to my work on Wood Thrush, I am also collaborating with colleagues to develop full-annual-cycle models to quantify the environmental drivers of migration survival in several species of Wood Warblers and to inform conservation planning efforts for Rusty Blackbirds, the fastest declining songbird in North America. In collaboration with the Cornell Lab of Ornithology, I am also beginning a project to investigate the declines of the eastern population of Painted Buntings. We have recently obtained pilot funding to determine the migration routes and winter distribution of this species by deploying approximatly 100 light-level geolocators across the entire breeding range of this population. This work is expected to begin in spring 2017. Data from these tags will ultimately be used to link partners working on Painted Bunting conservation throughout the annual cycle and to guide future monitoring, management, and outreach activities. Stay tuned for more!

Long-distance dispersal and the scale of source-sink dynamics

Immigration and emigration (collectively termed dispersal) are two of the primary processes that determine population dynamics. The consequences of these movements are poorly understood due to the logistical challenges associated with tracking dispersing in the field. A major theme of my research has been developing novel tracking and analytical tools to study the individual- and population-level consequences of long-distance dispersal.

I use observational and experimental fieldwork, stable isotope analysis, and statistical modeling to study the causes and consequences of long-distance dispersal in migratory songbirds. My work on American Redstarts produced a number of novel insights, including: 1) climatic conditions during breeding and winter periods drive long-distance dispersal (Rushing et al., 2015b); 2) reproductive success is influenced by winter habitat quality but not long-distance dispersal (Rushing et al., 2016b); and 3) philopatric and dispersing individuals use different cues to select breeding territories (Rushing et al., 2015c).

Currently, I am combining stable isotope analysis with integrated population modeling to understand the scale and drivers of immigration in Wood Thrush breeding populations. This approach facilitates the quantification of immigration at different spatial scales, allowing me to: 1) Distinguish local vs. regional immigrants in local populations; 2) quantify the demographic and environmental processes that drive local vs. regional immigration; and 3) quantify the contribution of immigrants to local population dynamics and determine the scale of source-sink dynamics. As part of a collaboration between scientists at the Smithsonian and Oregon State University, results from this work will be combined with other data on sources of origin (e.g. genetics) and modeling frameworks (e.g. occupancy modeling) to provide deeper insights into population dynamics and compare methods for quantifying vital rates.